We consider three major drivers as to the cause of this fortuitous development. First, the efficacy of pre-existing biological control organisms may have improved such that the contemporary CAS population is being controlled at levels that did not occur during the previous years. Second, a chance arrival and establishment of a new biological control organism may have occurred on Guam in the recent past and this new CAS enemy has not been identified to date. Third, the persisting living trees may have a degree of genetic resistance to CAS herbivory that the rest of the 2003 population did not have and the 96% mortality between 2003 and 2020 (Marler and Krishnapillai 2020) has selectively removed the less resistant genotypes from the population. These three broad drivers of population dynamics may not be mutually exclusive, but developing a greater understanding of their roles in the persistence of C. micronesica will improve future species recovery efforts.

We believe that a recent improvement in biological control efficacy of CAS is the most likely cause of the recent improvements in plant health and the details of this improved biocontrol have not been adequately studied. However, we also consider how the current trees’ survival was accomplished during the early years, and it may be through some form of resistance (antixenosis, antibiosis, or tolerance, nutritional status) (Emden 2002). As a result, the causes are not understood without more dedicated research. We will discuss evidence that suggests where plant resistance or biocontrol may or may not play a role in plant recovery.

• First, if the 2024 C. micronesica trees that appear to have improved in health status comprised the small percentage of 2003 trees endowed with the phene of greater constitutive genetic resistance to CAS in terms of antixenosis (plant surface, texture or volatiles etc., that make the plant unattractive) or antibiosis (negative effects on life cycle of the herbivore; for example, through chemical defences), these individual trees would likely have shown greater health and less CAS herbivory throughout the entire time span since 2003. This was not the case. After the CAS population was ubiquitous amongst all Guam habitats by 2006-2007, there were no C. micronesica trees that exhibited consistently greater health than the remaining sympatric trees (Marler and Lawrence 2012, Marler and Krishnapillai 2020). On the contrary, every tree within a habitat exhibited similar levels of CAS infestation and declines in health during every post-invasion year for more than a decade. If constitutive or induced genetic resistance is responsible for the recent increases in health status of the persisting trees, we believe that genetic resistance would have expressed itself during the past 20 years and it would not begin to express itself after almost two decades of scale herbivory. Resistance through tolerance or nutritional status cannot be ruled out, however. Plants that were able to thrive throughout the last 20 years may have had better starch storage or other abilities to recover from the herbivory through the years of continuous attacks by CAS.
• Second, some of the 2024 trees supported by a crown of healthy-looking leaves with minimal CAS herbivory have produced basal leaves from suckers or pups. These leaves near the soil surface invariably become heavily infested with CAS. An individual plant with genetic antibiosis chemical resistance to CAS would not likely exhibit one level of resistance at the 3-4 m stratum and a different level of resistance at the 0-1 m stratum. These observations indicate that differential levels of biological control as defined by vertical stratification (see section on regeneration below) (Marler 2013, Marler et al. 2013) are causal of the strata heterogeneity in CAS herbivory within individual trees. However, these observations do not rule out antixenosis resistance if leaves or plant parts of different maturities have different features (toughness, chemistry) that affect establishment by CAS crawlers.
• Third, some of the 2024 trees contain multiple branches and commonly one can find trees with heavy CAS infestation in one area of the canopy and minimal infestation in a second area of the same canopy. The most parsimonious explanation for this spatial heterogeneity of CAS infestation level within the same tree’s canopy is spatial heterogeneity of biological control, as every part of one tree’s canopy would likely contain the same level of genetic resistance to CAS. The question why some recovered trees have partial canopy heavily infested may be due to spatial conditions where biocontrol is restricted. As above, we cannot rule out that some differences in the location of the leaves (e.g. sun/shade) may affect leaf chemistry or physical traits that may play a role in CAS establishment.
• Fourth, regeneration and recruitment have been severely reduced by the chronic CAS herbivory, and very few ovules that are fertilised during the pollination phase have grown to maturity due to direct CAS herbivory of the developing seeds. In some of the habitats where mature seed production has happened in recent years, the germinating seedlings readily become heavily infested with lethal levels of CAS shortly after emergence of the first leaf. If Guam’s persisting mature trees are a result of their greater genetic antibiosis chemical resistance to CAS, then the pollen parent and ovulate parent of every contemporary seedling is endowed with this genetic resistance. Yet 100% of the newly-emerging seedlings continue to be rapidly killed by immediate CAS herbivory. There is the possiblity that antibiosis or antixenosis resistance, via chemical or physical properties, differs with maturity of leaves or plant stage as found in some other plant groups (such as Souza et al. (2021)). In cycads, for example, a chemical produced by cyanobacteria or the strain of cyanobacteria in the coralloid roots of mature plants may affect CAS, as has been demonstrated for some plants with symbiotic N-fixing rhizobia (Godschalx et al. 2015, Iriart et al. 2024). However, based on the results of a study that examined biocontrol of CAS, the stratification/height of seedling leaves was a factor, with biocontrol active only in higher structures of the plant (Marler et al. 2013).